Abstract
Spatial attention is thought to play a critical role in feature binding. However, there are often multiple objects or locations of interest in our environment, and we need to switch or split attention between them. How do these "attentionally ambiguous" scenarios affect the binding process? Here we show that binding is altered in both cases, but switching vs splitting attention result in different types of feature binding errors. We used a continuous report color perception task where subjects were pre-cued with a location(s) to attend, and then presented with an array of four colored squares. The task was to report the color at the attended location by clicking on a colorwheel. In the Attention-Switching experiments, we cued subjects to shift attention to a second location before the colors appeared. When the color stimuli were presented shortly after the switch cue, "swapping" errors were prominent, with subjects sometimes misreporting the color that appeared at the initially cued location. After longer delays (>500ms), the swapping errors diminished back to chance levels. In the Attention-Splitting experiments, subjects received two simultaneous cues and were instructed to attend to both locations; only after the colors appeared were they instructed which location to report. Here we found no evidence of swapping errors, but we instead found "mixing" errors: subjects tended to report a color that was a subtle mix of the target color and the color at the other attended location. Interestingly, in another version we varied the distance in color space between the attended colors: with large differences in color we replicated the mixing result, but with small color differences we found repulsion. We suggest that "swapping" and "mixing/repulsion" errors stem from different attentional mechanisms: incomplete updating of attention results in swapping errors, whereas mixing/repulsion errors occur when two locations are simultaneously sharing attentional resources.
Meeting abstract presented at VSS 2014